fined-tuned-bart / README.md
ppiiesle3y's picture
Update README.md
95265c8
metadata
language:
  - en
tags:
  - summarization
license: mit
datasets:
  - multi_news
model-index:
  - name: ppiiesle3y/fined-tuned-bart
    results:
      - task:
          type: summarization
          name: Summarization
        dataset:
          name: multi_news
          type: multi_news
          split: train
        metrics:
          - name: ROUGE-1
            type: rouge
            value: 43.7065
            verified: true
          - name: ROUGE-2
            type: rouge
            value: 16.5533
            verified: true
          - name: ROUGE-L
            type: rouge
            value: 24.7588
            verified: true
          - name: ROUGE-LSUM
            type: rouge
            value: 37.7586
            verified: true
          - name: loss
            type: loss
            value: 2.00663
            verified: true
          - name: gen_len
            type: gen_len
            value: 129.1379
            verified: true

TL;DR AT2 Applied Natural Language Processing Assignment

PROJECT OBJECTIVES

This project aims to use NLP technology to summarise longer passages of text into succinct and accurate summations.

PROJECT OUTCOMES AND INSIGHTS

The expected outcomes from the project is a model that is able to intake a larger body of text and provide a shortened summary that is both succinct and accurate. This will benefit most human readers by making it more efficient gain understanding from written text. Applications for this technology include as a study aide, for people in roles where they are required to quickly assess documents such as book publishers reading through manuscripts to assess if they are fit for publishing or script readers etc. The most significant impact this project has is to increase information assimilation in a compressed timeframe, thus saving time.